Developing a surveillance ordinance in Cambridge

On Wednesday night, Cambridge’s Public Safety Committee held a hearing to discuss a question that hangs over cities across the country: who gets to control whether and how new technology is used for surveillance? Will police be permitted to acquire new surveillance technologies without notifying anyone?

The consensus at City Hall was no. The City Manager, other City officials, and the public expressed an eagerness develop a surveillance ordinance that will require democratic, public approval of any technology that could be used for surveillance. I was encouraged by the productive and agreeable conversation we had, and look forward to seeing what comes next.

The challenge ahead is to develop a surveillance ordinance that appropriately accounts for the complex surveillance risks of new technology, and that is future-proofed against those to come. A few of my recommendations—presented during the hearing—are below:

With technology developing at such a rapid pace, cities must develop policies that comprehensively account for what surveillance means today and how it is conducted. The first question a city must ask is what technology should be classified as related to surveillance. Traditional conceptions about privacy lead people to focus on the tangible forms of being watched: cameras, sensors, and even police officers. But as with most things, the nature of surveillance has evolved significantly over the past several years. Today, many forms of surveillance occur through software that analyzes the data collected by surveillance equipment or scraped from social media feeds.

Failing to account for today’s realities of surveillance — and the forms of surveillance most likely to become possible in the future — will render a surveillance ordinance impotent. Let’s learn from Seattle, which today has perhaps the nation’s most democratic and privacy-protective policies regarding the adoption of new technology. In its first implementation of the bill, in 2013, Seattle limited its mandate of public input and city council approval to purchases of new equipment. Last year, the Seattle Police were exposed as using Geofeedia, a service that collects and analyzes social media data. This was a notable expansion of police surveillance — but also acceptable within the bounds of Seattle’s 2013 surveillance ordinance. Geofeedia is software, not equipment, and so public oversight did not apply.

It is also essential that public oversight be applied to the policies that govern surveillance technology, in addition to whether that technology should be used at all. Without appropriate controls, technologies intended for one purpose can be twisted for another. Body cameras have been adopted by countless cities in the past several years as a tool for police accountability: many hoped that police, knowing their behavior was being recorded, would act more respectfully toward the public (and those that acted inappropriately would be held accountable). But as police departments adopted body cameras, they failed to implement policies that enforced the availability of footage as a resource for police oversight and restricted how recordings could be used for other purposes. If police are given free rein to use and analyze video footage, the presence of widespread body cameras could massively expand surveillance.

These issues — evolving definitions of surveillance technology, and the policies that govern it — are merging in troubling ways as body camera manufacturers develop facial recognition software, which police could use to track people’s movements and identify individuals in crowds. This capability could be deployed to every police department that has body cameras through a software update, much like you update your iPhone. Future-proofing a surveillance ordinance thus requires comprehensively enforcing public oversight of software updates that meaningfully alter a technology’s capabilities for surveillance. Otherwise, a technology approved in one form (or perhaps not even considered surveillance technology when purchased) can evolve into a tool for surveillance as its software is updated.

This is not a debate about being for or against innovation — it is about facilitating the innovation that city residents want. For there is more to innovation than adopting new technology. Innovation also means adapting policies and practices to better achieve your values.

By working to enforce stronger public oversight of surveillance technology, Cambridge can help ensure that new technology is deployed in accordance with its residents’ values and improves daily life.